Transit Ridership Modeling at the Bus Stop Level: Comparison of Approaches Focusing on Count and Spatially Dependent Data

Abstract Boarding and alighting modeling at the bus stop level is an important tool for operational planning of public transport systems, in addition to contributing transit-oriented development. The interest variables, this case, present two particularities that strongly influence performance proposed estimates: they demonstrate spatial dependence are count data. Moreover, most cases, these data not easy collect. Thus, study proposes a comparison approaches transit ridership level, applying linear, Poisson, Geographically Weighted Poisson (GWPR) regressions, as well Universal Kriging (UK), boarding along line city São Paulo, Brazil. results from goodness-of-fit measures confirmed assumption adding asymmetry autocorrelation, isolated together, transportation demand modeling, contributes gradual improvement estimates, highlighting GWPR UK estimation techniques. spatially varying relationships between variables (boardings alightings) their predictors (land use system features around stops), shown study, may support land policies toward In addition, by using approach with little information, good achieved proved satisfactory can be done regions where there lack travel data, case emerging countries.